North Pacific Observer Program Reporting

July 01, 2015

Alaska Fisheries Science Center Quarterly Report

Feature Story |

Alaska

North_Pacific_Groundfish_and_Halibut_Observer_Program1.jpg

At the June 2015 meeting of the North Pacific Fishery Management Council in Sitka, Alaska, the Fisheries Monitoring and Analysis Division (FMA) presented the North Pacific Groundfish and Halibut Observer Program 2014 Annual Report to the Council.

The Annual Report, which is drafted jointly by staff from the Alaska Fisheries Science Center and the Alaska Regional Office, provides information, analysis, and recommendations based on deployment of observers in the North Pacific Groundfish and Halibut Observer Program. The Observer Program provides the regulatory framework for NOAA Fisheries certified observers to obtain information necessary to conserve and manage the groundfish and halibut fisheries in the Gulf of Alaska and the Bering Sea and Aleutian Islands management areas. Data collected by trained, independent observers are a cornerstone of management of the Federal fisheries off Alaska. These data are needed by the Council and NOAA Fisheries to comply with the Magnuson-Stevens Fishery Conservation and Management Act, the Marine Mammal Protection Act, the Endangered Species Act, and other applicable Federal laws and treaties.

Each year NOAA Fisheries releases an Annual Deployment Plan (ADP), the most recent of which was released in December 2014, which describes how NOAA Fisheries plans to deploy observers to vessels and processors in the partial observer coverage category in the upcoming year. The following year, the agency provides an Annual Report with descriptive information and scientifically evaluates the deployment of observers. The Annual Deployment Plan and Annual Report process provide information to assess whether the objectives of Observer Program restructuring have been met and a process to make recommendations to improve implementation of the program to further these objectives. The 2014 Annual Report provides information and recommendations based on deployment of observers in 2014.

In addition to scientifically evaluating the deployment of observers in the partial observer coverage category, the Annual Report includes information on fees and budget for the partial coverage category (Chapter 2); descriptive information on the number of vessels and trips, and catch and discards by FMP area, and a summary of observer trainings and briefings (Chapter 4); compliance and enforcement (Chapter 5); outreach (Chapter 6); and NOAA Fisheries recommendations (Chapter 7).

Observer Coverage Categories and Coverage Levels

Under the current Observer Program, all vessels and processors in the groundfish and halibut fisheries off Alaska are assigned to one of two observer coverage categories: 1) a full coverage category or 2) a partial coverage category.
The full coverage category includes:

Catcher-Processors

Motherships.
Catcher vessels while participating in programs that have transferable prohibited species catch (PSC) allocations as part of a catch share program.
Inshore processors when receiving or processing Bering Sea pollock.

The Partial Observer Coverage Category Includes

Catcher vessels designated on a Federal Fisheries Permit when directed fishing for groundfish in federally managed or parallel fisheries, except those in the full coverage category.
Catcher vessels when fishing for halibut individual fishing quota (IFQ) or sablefish IFQ (there are no PSC limits for these fisheries).
Catcher vessels when fishing for Community Development Quota (CDQ) halibut, fixed gear sablefish CDQ, or groundfish CDQ using pot or jig gear (because any halibut discarded in these CDQ fisheries does not accrue against the CDQ group’s transferable halibut PSC allocation).
Catcher-processors that meet criteria that allow assignment to the partial coverage category.
Shoreside or stationary floating processors except those in the full coverage category.

Under the 2014 Annual Deployment Plan , the partial coverage category consisted of vessels in three “strata” (statistical subgroups) or “pools” with differing requirements:

No Selection Pool

This category applied to all vessels less than 40 ft length overall (LOA) and catcher vessels fishing with jig gear (which includes handline, jig, troll, and dinglebar troll gear). Inclusion in this pool is re-evaluated each year in the Annual Deployment Plan and may change in the future. Eligible landings from vessels in the no selection pool are included in the observer fee assessment.

Vessel Selection Pool¹

This category applied to catcher vessels fishing with hook-and-line and pot gear that are greater than or equal to 40 ft and less than 57.5 ft LOA. Vessel owners or operators in this pool were not required to log trips into the Observer Declare and Deploy System. However, a subset of vessels, randomly selected by NOAA Fisheries, were required to take observers for every groundfish or halibut fishing trip that occurred during a specified 2-month period. Owners of selected vessels were contacted by NOAA Fisheries at least 30 days in advance of the 2-month period.

Trip Selection Pool

This category applies to all catcher vessels of any length fishing with trawl gear; to hook-and-line and pot gear vessels that are greater than or equal to 57.5 ft. LOA; and to the small catcher-processors eligible to be placed in partial coverage. Owners or operators of vessels in this pool are required to log each fishing trip into the ODDS system. Upon logging a trip, the vessel owner or operator is immediately informed if the trip has been randomly selected for observer coverage. If the logged fishing trip is selected, then the vessel must take an observer on that trip. The observer will be provided by a NOAA Fisheries contractor. Vessel owners or operators in this pool must log fishing trips at least 72 hours before anticipated departure.

NOAA Fisheries Set the Final Deployment Rates for 2014

12% of vessels for the vessel selection pool, and
16% of trips for the trip selection pool.

( ¹The vessel selection pool was discontinued at the end of 2014 due to concerns about the quality of observer data from vessels in this pool. It was difficult for NOAA Fisheries to accurately project the list of vessels that would fish in the vessel selection pool in each selection period, which made it difficult to randomly select vessels for observer coverage. In addition, the large number of conditional releases created concerns about bias in the data collected from vessels in this pool. These concerns are described in more detail in the 2013 Annual Report. )

Key Results of the 2014 Observer Program Deployment

The following is a summary of some of the accomplishments of the Observer Program in 2014. Note that some of these accomplishments combine statistics from both full and partial coverage categories.

A total of 436 individual observers were trained, briefed, and equipped for deployment to vessels and processing facilities operating in the Bering Sea and Gulf of Alaska groundfish fisheries.
The Observer Declare and Deploy System performed as expected with only one service interruption and 4,687 trips logged.
Rates of observer deployment at-sea were within expected ranges.
Observers collected data onboard 417 fixed-gear and trawl vessels and at 14 processing facilities for a total of 44,178 observer days.
With few exceptions, the observer provider company AIS Inc. successfully deployed observers from 30 ports for 4,368 days at sea with a minimal number of trips released from coverage.
There were 97 debriefings completed in Anchorage by 3 FMA staff and 572 debriefings in Seattle completed by 21 FMA staff.
NOAA Fisheries held outreach events in Seattle, Petersburg, Sitka, Juneau, Kodiak, and Homer to maintain a dialogue with industry members and inform them about changes to the program, vessel responsibilities, and the objectives of quality collection of data and management.

Did We Meet Anticipated Deployment Goals

Number of total vessels (V), sampled vessels (v), total trips (N), sampled trips (n) for each observer deployment stratum in 2014. Totals are unique vessels. Expected coverages are in percent for trip selection and number of vessels for vessel selection. TS= Trip selection, VS= Vessel selection, ZS= Zero selection, EM= Electronic Monitoring. Vessel selection, ZS= Zero selection, EM= Electronic Monitoring.

The Observer Program met expected rates of coverage for the full-coverage regulatory and full-coverage voluntary strata, the trip selection stratum, four of six time-periods within vessel selection, and the partial coverage no selection.

Observer coverage was higher than the expected 12% selection rate in two of the six time periods within vessel selection. Vessels were selected for sampling based on whether they fished in 2013. This resulted in a discrepancy between the sampling list and the list of vessels that actually fished. In addition, the unpredictability in the number of vessels that would be granted conditional releases meant that NOAA Fisheries “oversampled” in some of the vessel selection time periods. These problems were also highlighted in the 2013 Annual Report and were part of the rationale for moving all vessels to the trip selection method in 2015.

Random selection of trips in the trip selection stratum is facilitated by the ODDS system. Users of the system are given flexibility to accommodate their fishing operations; up to three trips may be logged in advance of fishing, and trips can be canceled to accommodate changing plans. Once a trip has been completed, logged trips must be closed by a vessel operator.

If a trip is selected for observer coverage and canceled by the user, then the vessel's next logged trip is automatically selected for coverage. The "inherited" trips preserve the number of selected trips in the year, but cannot prevent the delay of selected trips during the year, which resulted in temporal bias.

In 2014, ODDS users canceled trips that had been selected for coverage at nearly four times the rate of unselected trips. Since only canceled trips that had been originally selected are preserved, the final selection rate in ODDS was higher than if selected trips had not been disproportionately canceled.

Was the Coverage Representative

North_Pacific_Groundfish_and_Halibut_Observer_Program4.jpg

An observer sampling on a trawl-catcher vessel.

NOAA Fisheries evaluated the possibility for temporal bias in the trip selection stratum. Although coverage rates were lower than expected at the beginning of the year, the final coverage rate was within expected ranges. In 2014, the spatial distribution of observer coverage in trip selection was as would be expected under a random sample of trips. In vessel selection, however, there were more observed vessels in certain NOAA Fisheries reporting areas than would be expected under random deployment. This result highlights the difficulty in obtaining an adequate sampling frame in vessel selection.

NOAA Fisheries expanded the comparison of trip metrics between various categories of vessels relative to the analysis conducted in the 2013 Annual Report. In both the 2013 version and the 2014 Annual Report, NOAA Fisheries compared trip duration (number of days), number of NOAA Fisheries areas visited during a trip, landed catch weight, species diversity (the number of different species in the landed catch), and the proportion of landed catch that was due to the predominant species in the catch (the “purity” of the catch). For 2014, NOAA Fisheries added comparisons of vessel length to the trip metrics and performed analyses using permutation tests instead of visual inspection of histograms.

Comparison of Tender Trips and Non-Tender Trips

Vessels that delivered to a tender were 11.5% shorter in length, fished 29.1% longer in duration, and had catch that was 1.3% less “pure” than vessels that did not deliver to a tender.

Comparison of Observed and Unobserved Trips Delivered to Tenders

The analysis found no differences in NOAA Fisheries areas visited during a trip, trip duration, the total weight of landed catch, or the number of species in the landed catch. The permutation tests did, however, indicate a difference in vessel length and the proportion of the predominant species in the landed catch for observed and unobserved vessels delivering to tenders. Observed vessels delivering to tenders were 8.8% shorter than unobserved vessels delivering to tenders. The landed catch by observed vessels delivering to tenders was 6% less “purely the predominant species” than landed catch by unobserved vessels delivering to tenders.

Comparison of Observed Unobserved Trips Delivered Shoreside

Trip selection

Hook-and-line vessels that were observed landed 14.4% less catch and 9.1% more species than unobserved vessels. Trawl vessels that were observed fished in 4.2% fewer areas and were 8.4% shorter in duration than unobserved vessels. There were no differences between observed and unobserved trips for vessels that fished pot gear. Taken together, there is evidence of an observer effect in trip selection hook-and-line and trawl gear.

Vessel selection

Six trip metrics were evaluated for each of the 6 time periods (36 comparisons). Of the 36 tests, 18 showed significant differences between observed and unobserved trips providing evidence of an observer effect in vessel selection.

Based on the 2014 Annual Report and recommendations made by NOAA Fisheries, the Council provided a number of recommendations for consideration in the development of the 2016 Annual Deployment Plan and future annual reports, including:

Provide additional information on observer rates and percent coverage by gear type, in addition to numbers of trips by coverage category (full or partial coverage).
Determine best approaches to a trip identifier tied to landings data to provide a linkage between ODDS and eLandings and improve data analysis, including those trips delivered to a tender.
Evaluate and suggest modifications to ODDS to reduce temporal bias associated with the policy of allowing trip cancelation and logging multiple trips prior to departure.
Develop tools to evaluate both the reliability of the data and deployment performance.
Assess inefficiencies in the program and ways to achieve cost efficiencies in the partial coverage category within the existing 5‐year contract.
Provide information about the availability of fixed-gear lead level 2 observers.
Add quantitative measures in the enforcement section of the report, especially in relation to trends by incident type.
Define strata to deploy observers by gear (longline, pot, and trawl gear) and FMP area and, if necessary, consider operational sector (CV vs CP).

NOAA Fisheries is in the process of considering these recommendations and, to the extent possible, incorporating them into the Draft 2016 Annual Deployment Plan, which will be presented to the Council at the October 2015 meeting. This iterative process of analyzing the previous year’s deployment performance provides the Agency and the Council with flexibility to amend and improve the Observer Program on an annual basis.